摘要
基于机电阻抗(EMI)方法的健康监测技术近年来迅速发展,但温度对该方法的影响使得EMI技术的发展受到了一定的阻碍。此研究以实验为基础分析了温度对基于EMI的损伤检测技术的影响,随后分别采用有效频率偏移方法和神经网络方法对EMI损伤检测技术做温度补偿,并对比分析了两者的补偿效果。实验研究表明温度的变化会引起EMI方法中导纳曲线的偏移,引起RMSD损伤指标的变化,从而影响对损伤程度的分析判断。通过有效频率偏移方法及神经网络方法作温度补偿后,均得到了比较理想的补偿效果,温度影响明显减小。通过对比得出神经网络方法的补偿效果更加明显。
Electro-mechanics impedance （ EMI ） based structure health monitoring has a rapid development in recent years. But the temperature effect on EMI technique hinders the development of this technique. An experiment was performed to research the temperature effect on EMI based damage detection technology. Both the effective frequency shift （ EFS） method and artificial neural network （ ANN） method were adopted to compensate the temperature effect, and the compensation result was analyzed to compare the efficiency between two methods. The experimental result shows that the admittance curve will shift with the increase of temperature in EMI method, the RMSD index is changed because of the temperature effect. So this effect can mislead the judgment of the damage detection. Good results are both obtained after the temperature compensation by EFS and ANN methods, and the ANN method is more effective by comparing of compensation result.
出处
《土木工程与管理学报》
2014年第3期7-11,33,共6页
Journal of Civil Engineering and Management
基金
国家重点基础研究发展计划(973计划)(2011CB013804)
国家自然科学基金(51278215)
关键词
机电阻抗
损伤检测
温度补偿
神经网络
electromechanical impedance
damage detection
temperature compensation
artificial neural network
